Psychotic depression (PD) is a severe disorder that is present in 20% of mixed-age inpatients and in more than 40% of elderly inpatients hospitalized for the treatment of major depression. Acutely, PD is associated with a greater level of depression severity, functional impairment, and resistance to antidepressant treatment than nonpsychotic depression. Longitudinally, PD is associated with a greater likelihood of recurrence, a shorter symptom-free interval, greater disability, and an increased long-term risk of suicide. Among elderly patients, PD has also been associated with a high one year mortality rate. Although electroconvulsive therapy (ECT) is a highly effective treatment for PD, pharmacotherapy is more easily administered and generally preferred by many patients and psychiatrists. Furthermore, ECT is less available to, and less frequently used by minority ethnic groups, patients with low incomes, and those residing in rural areas. In the early 1980's a landmark NIMH randomized controlled trial demonstrated that a combination of high doses of both a tertiary amine tricyclic and a conventional antipsychotic was more efficacious than monotherapy with either medication. However, this combination has been associated with a high incidence of significant side effects, particularly in elderly patients. This collaborative U01 project builds on findings from the four sites principal investigators and others. Based on preliminary studies suggesting that sertraline, a serotonin-reuptake-inhibitor (SSRI) and olanzapine, an atypical antipsychotic, may be effective in the acute treatment of PD, this study uses a two-armed randomized double-blind parallel-group design to compare the efficacay and tolerability of combining these two medications to olanzapine alone. By blocking recruitment of subjects age >60 on a 1:1 basis the study will determine the relationship between older age and remission, time to remission and tolerability. A 12-week acute phase efficacy study is followed by a 20-week continuation trial during which subjects who achieved partial remission enter a placebo-controlled augmentation trial that is based on their initial assignment. Remitted subjects are also followed during stabilization to determine the stability of remission and predictors of clinical and functional deterioration.